US9317279B2

US9317279B2 - Virtual machine block substitution

Info

Publication number: US9317279B2
Application number: US14/298,902
Authority: US
Inventors: Robert Fries; Srivatsan Parthasarathy
Original assignee: Microsoft Technology Licensing LLC
Current assignee: Microsoft Technology Licensing LLC
Priority date: 2011-06-29
Filing date: 2014-06-07
Publication date: 2016-04-19
Anticipated expiration: 2031-06-29
Also published as: US8819660B2; US20130007733A1; US20140289720A1

Abstract

Techniques are described to allow substation of blocks into a virtual machine image of an executing virtual machine. A patch may be applied to a first virtual machine image. One or more blocks modified by the patch are identified and copied to a host having a virtual machine executing from a virtual machine image that is based on the first virtual machine image (prior to the patch). While the virtual machine is executing, the substitution blocks are applied (supersede) corresponding original blocks in the virtual machine image. A guest operating system in the virtual machine may either begin using an executable file in a substitution block. In addition, the guest operating system may be forced to flush from memory and cache executable code replaced by the substitution block.

Description

CROSS-REFERENCE TO RELATED APPLICATION

This application is a Continuation of application Ser. No. 13/171,443, filed Jun. 29, 2011, allowed and to be issued, and hereby incorporated in original form by reference.

BACKGROUND

Data centers, compute clouds, and other networks of cooperating hosts often use virtual machines (VMs) for management and efficiency benefits. In some cases, a virtual machine image such as a “golden image” may be duplicated and used in multiple virtual machines. That is, a root or base virtual machine image may exist at a management server, and virtual machines on hosts may be running copies of the base virtual machine image.

At times, it may be desirable to update or patch the software such as applications and operating systems on virtual machines. Often, the same patch must be applied to all of the virtual machines that are based on a same virtual machine image. That is, virtual machines using copies of a base virtual machine image may need a same patch or update. Typically, this has involved, for running virtual machines, copying a patch, installing the patch, and restarting the virtual machine and its guest operating system. However, restarting a virtual machine may present problems. Particularly in environments of distributed or highly interconnected applications, or with applications that require minimal downtime, restarting may disrupt a service or affect a Service Level Agreement. Even in cases where there is only virtual machine to be updated, it may nonetheless be inconvenient to restart that virtual machine.

Techniques related to updating virtual machine images of live virtual machines are discussed below.

SUMMARY

The following summary is included only to introduce some concepts discussed in the Detailed Description below. This summary is not comprehensive and is not intended to delineate the scope of the claimed subject matter, which is set forth by the claims presented at the end.

Many of the attendant features will be explained below with reference to the following detailed description considered in connection with the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

The present description will be better understood from the following detailed description read in light of the accompanying drawings, wherein like reference numerals are used to designate like parts in the accompanying description.

FIG. 1 shows an example virtualization layer.

FIG. 2 shows processes and interactions of virtualization layer in relation to virtual machines and virtual machine images.

FIG. 3 shows a process for substituting blocks into an executing virtual machine.

FIG. 4 shows a virtual machine image being patched.

FIG. 5 shows substitute or modified blocks being applied to a target virtual machine image of an executing virtual machine.

FIG. 6 shows a detailed view of physical memory and storage when a substitute block is loaded.

DETAILED DESCRIPTION

Virtualization Overview

FIG. 1 shows an example virtualization layer 100. A computer 102 has hardware 104, including a central processing unit (CPU) 106, memory 108, a network interface 110, non-volatile storage 112, and other components not shown, such as a bus, a display adapter, etc. The virtualization layer 100 manages and facilitates execution of virtual machines 114. Although not shown in FIG. 1, each virtual machine 114 typically has an associated virtual disk image and a guest operating system. For brevity, the operating system and perhaps application software of a virtual machine 114 will sometimes be referred to as a guest, which is stored and executed from the virtual disk image associated with the virtual machine 114.

The virtualization layer 100 may be of any variety of known or future implementations, such as Hyper-V Server™, VMWare ESX Server™, Xen, Oracle VM™, etc. The architecture of the virtualization layer may a hosted type, with a virtual machine monitor (VMM) running on a host operating system, or a bare-metal type with a hypervisor or the like running directly on the hardware 104 of the computer 102. As used herein, the term “virtual machine” refers to a system-type virtual machine that simulates any specific hardware architecture (e.g., x86) able to run native code for that hardware architecture; to the guest, the virtual machine may be nearly indistinguishable from a hardware machine. Virtual machines discussed herein are not abstract or process-type virtual machines such as Java Virtual Machines.

The virtualization layer 100 performs the basic function of managing the virtual machines 114 and sharing of the hardware 104 by both itself and the virtual machines 114. Any of a variety of techniques may be used to isolate the virtual machines 114 from the hardware 104. In one embodiment, the virtualization layer may provide different isolated environments (i.e., partitions or domains) which correspond to virtual machines 114. Some of the virtualization layer 100 such as shared virtual device drivers, inter virtual machine communication facilities, and virtual machine management APIs (application programming interfaces), may run in a special privileged partition or domain, allowing for a compact and efficient hypervisor. In other embodiments, functionality for virtual machine management and coherent sharing of the hardware 104 may reside in a monolithic on-the-metal hypervisor.

FIG. 2 shows processes and interactions of virtualization layer 100 in relation to virtual machines 114 and virtual machine images 140. The virtualization layer 100 performs a process 142 of starting and executing a virtual machine 114, possibly according to corresponding virtual machine configuration parameters. When a virtual machine 114 (VM) is started, the virtualization layer identifies an associated virtual machine image 140. In practice, any virtual machine image 140 can be used by any virtual machine 114. The virtual machine image 140 may be a specially formatted file (e.g., a VHD) on a file system 141 of the virtualization layer 100. The virtualization layer 100 loads the identified virtual machine image 140. The started virtual machine 114 mounts and reads the virtual machine image 140, perhaps seeking a master boot record or other boot information, and boots a guest operating system which begins executing.

The virtualization layer 100 manages execution of the virtual machine 114, handling certain calls to the guest's kernel, hypercalls, etc., and coordinating the virtual machine 114's access to the underlying hardware 104. As the guest and its software run, the virtualization layer 100 may maintain state of the guest on the virtual disk image 140; when the guest, or an application run by the guest, writes data to “disk”, the virtualization layer 100 translates the data to the format of the virtual disk image 140 and writes to the image.

The virtualization layer 100 may perform a process 144 for shutting down the virtual machine 114. When an instruction is received to stop the virtual machine 114, the state of the virtual machine 114 and its guest is saved to the virtual disk image 140, and the executing virtual machine 114 process (or partition) is deleted. A specification of the virtual machine 114 may remain for a later restart of the virtual machine 114.

Live Block Substitution

FIG. 3 shows a process for substituting blocks into an executing virtual machine. At step 160 substitution blocks are identified. This may be accomplished in a number of ways. In one embodiment, a patch is applied to a first virtual machine image (e.g., a golden image). The patch may be applied by mounting the first virtual machine image as a file system and modifying files or other guest data stored in the virtual machine image. The patch may be applied by booting the virtual machine image in a virtual machine and executing a patching process (e.g., a Microsoft Installer™ file) to update an operating system or software installed thereon, including modifying a binary machine executable file. One or more blocks in the second (patched) virtual machine image may be identified by comparing the first and second virtual machine images block by block. If only a few select files containing binary machine executable instructions have been updated, it may be possible to identify the tainted block(s) (blocks containing the modified code) directly during the patching process. For example, a same block in the first and second virtual machine image (e.g. block N) may contain an original and updated file, respectively. By whatever means, it is assumed that at least one tainted block is identified in the second virtual machine image.

At step 162, the substitution blocks are copied to one or more target virtual machines. The substitution blocks may or may not be accompanied by a list identifying the substitution blocks. In one embodiment, the substitution blocks may be in the form of a difference disk or delta disk, details of which are known and described elsewhere. In another embodiment, the virtual machine images may be dynamic virtual disks that expand in size as needed.

At step 164 the blocks are applied at the hosts at which they are received. The substitution blocks are handled at the virtualization layer, which applies the substitution blocks to the virtual machine images thereon. For discussion, consider only a single host and a single substitution block. The host has a virtualization layer that manages execution of a virtual machine using a virtual machine image. The virtualization layer has a block-level driver that virtualizes access to the virtual machine image, providing block-level access to the image and causing the virtual machine image file to appear as a disk to the virtual machine. As such, the virtualization layer is able to transparently alter that “disk” without the need for awareness or cooperation of the virtual machine or the guest software executing in the virtual machine (although embodiments involving same are described later). The virtualization layer (or hypervisor) identifies a block in the target virtual machine image that will be replaced by the substitution block, possibly using a tainted block list received with the substitution block, or by comparing the substitution block with blocks in the target virtual machine image. The identified tainted block is then replaced with the substitution block, which becomes available to the still-executing associated virtual machine and guest software.

At step 166, operations below the virtualization layer may be performed. Note that in a case where the substitution block is known or assumed to contain read-only data, the substitution block may be copied directly and any substituted-in files may be simply used when they are accessed by the guest operating system. In another embodiment, the virtualization layer informs the guest operating system by a signal or message, possibly passed through a virtual machine bus or the like. For example, the hypervisor may send the virtual machine information about files in the substitute block that have been modified (possibly files with binary executable code). Or, by informing the operating system about the storage area of the block, and leaving it to the operating system to take any action, such as scanning the file system at file system locations corresponding to the block. In one embodiment, substitution blocks may contain customization information, which is specialized information to target or facilitation substitution. For example, a substation block may contain the name of the machine or may specify various machine configuration settings. When substitution occurs, the customization information is used to update a setting or state of the guest that is persisted in the corresponding virtual machine image. In this case, the substitution system has built-in understanding of the guest operating system, in contrast to embodiments where substitution is performed at the virtual hardware layer without regard for the guest operating system.

In one embodiment, the guest operating system may have a component such as a kernel module that responds to a signal from the hypervisor by taking various steps to assure that a file in the substitute block becomes available when the substitution block is copied into the virtual machine image. This may involve the guest operating system invalidating any paged-in executable code from the original (replace) block that is now out-of-date, invalidating or marking invalid any relevant file cache, and so forth. The guest operating system might also force a page fault to cause the file to be reloaded into memory (thus allowing for execution of code in the substitute block). The guest operating system may access a translation lookaside buffer to identify where the old code of a new file is located in physical memory, and update accordingly.

FIG. 4 shows a virtual machine image being patched. An original virtual machine image 190, such as a golden image, is to be patched. A patch 192 is applied to the original virtual machine image (VMI-A). As noted above, this may involve mounting the original virtual machine image 190 (e.g., with an administrative tool), or booting the original virtual machine image 190 in a virtual machine. The patch 192 is then applied, which may update files, registry settings, virtualization settings, executables (libraries, programs, etc.), and other known patchable objects. A number of original blocks 194 may be affected. Modified blocks 196 in the patched virtual machine image 198 (VMI-A′) are indicated by shading in FIG. 4. In one embodiment, a patch tool, virtual machine image tool (configured to read the format of the virtual machine images), or the like compares the original virtual machine image 190 and the patched virtual machine image 198 (e.g., by comparing block hashes, disk differencing, ec.). In another embodiment, blocks are tracked as they are updated and a list 200 of tainted blocks is thus accumulated. The substitute blocks 196 are then transmitted via a network to hosts where they will be applied.

FIG. 5 shows substitute or modified blocks 196 being applied to a target virtual machine image 220 of an executing virtual machine 222. A host 224 is hosting a hypervisor that runs the virtual machine 222 and manages access to the virtual machine image 220. The modified blocks 196 are received at the host 224 from a data center's virtual machine management software, for example. As the virtual machine 222 is executing, the modified blocks 196 are copied into the virtual machine image 220. In one embodiment, the tainted block list 200 is used to substitute in the modified blocks 196 for corresponding pre-existing outdated blocks. In another embodiment, if disk-differencing is used, the modified blocks 196 are effectively copied in by placing them in a difference-disk without explicitly overwriting old blocks. As the guest operating system 226 executes, it may implicitly or explicitly begin to access data, including possibly executable code, in the modified blocks 196. For example, an executable file 228 in a modified block 196 begins executing in the guest operating system 226 without restarting the guest operating system 226 or the virtual machine 222.

FIG. 6 shows a detailed view of physical memory and storage when a substitute block is loaded. A substitution block 196A may contain a portion of executable code 240A, for instance machine instructions for a processor of the relevant host. The executable code 240A originated from the patch 192. In some cases, it might be that the substitution block 196A is small and contains only patched/new executable code. In other cases, the substitution block 196A may contain some files or other data that has not changed (i.e., a file that was not changed by the patch 1 92). The hypervisor 242 updates the target virtual machine image 220 with the substitution block 196A, which replaces a previous block 244. The code 240A becomes available within the target virtual machine image 220 as code 240B (now a part of the virtual machine image 220 of the executing virtual machine 222).

In response to the insertion of the new substitution block 196A, or with some latency (for instance, when needed by the guest operating system 226), the code 240B is accessed by the guest operating system 226. At this time, the code 240B is paged into physical memory 246 by the virtualization layer, forming in-memory code 240C. The in-memory code 240C backs code 240D mapped in the virtual memory address space of the guest operating system 226.

It will be appreciated that any of the involved components may be modified or augmented to force out old code and force in new code. For example, the guest operating system 226 may be “enlightened” (provided with a virtualization-aware element) that flushes any affected caches, forces relevant pages to be faulted (and therefore re-read, thus loading code 240B into physical memory 246 as code 240C), etc. In another embodiment, the virtualization layer itself might directly update the physical memory of the guest operating system 226 with new code, possibly by scanning physical memory 246 to recognize the old code to be replaced. For instance, an appropriately enlightened operating system may float up to the hypervisor information about the relevant files and or memory. It is important to recognize that when the substitution block 196A is added the virtual machine may have an execution state that is inconsistent with the virtual machine image 220; code may still reside in memory that has been replaced by the substitution block 196A. A number of approaches are discussed below to address this possibility.

One approach for dealing with substituted-in code replacing code that has already been executed is to keep a pedigree for every memory-paged file. That is, the hypervisor tracks which disk block each file came from. With this approach, page table entries are invalidated, and the file cache may need to be cleared. In another embodiment, a page is taken, a page fault is forced, the page tables is searched, and the appropriate pages are faulted in, and the page tables are updated (possibly using a hook in the hypervisor). Another approach may involve the hypervisor passing to the guest a list of updated (invalid files), which the guest uses to refresh accordingly. In another embodiment, a file-to-block mapping is used to identify updated files, with appropriate memory fixes then performed.

To elaborate, consider that hypervisors typically provide a block device layer, and the guest operating system a file system layer. If the guest executes a binary file, then certain sections are mapped to physical pages; there is a page layer, which is backed by files (sometimes a swap file), backed by or directly mapped onto blocks. A file such as the text section of a binary file, may be loaded to a page.

While updates to a executing virtual machine have been mentioned, another embodiment may pause or suspend a virtual machine when a new block is substituted in. Although it may be convenient to use blocks that match a block-size used by the hypervisor performing the substitution, this is not required; smaller or larger blocks may be used. Moreover, although it may in some cases be desirable to avoid restarting virtual machines and guest operating systems, such may still be performed after a live block substitution, for example to resolve errors or to handle irreconcilable memory-disk issues.

Without limiting the embodiments, it should be noted that the live-patching technique discussed above may be used without regard for the underlying operating system. In addition, patching may be performed outside the ordinary mechanisms of virtual machine control and ownership. Specifically, a cloud hoster or data center operator can apply a patch to virtual machines without requiring guest or virtual machine credentials, which may be convenient for updating cloud fabric code that is embedded in a virtual machine. Furthermore, it may be possible to update software on a virtual machine without having to restart the virtual machine or the guest operating system. Some embodiments may be more readily implemented when update blocks contain read-only files.

CONCLUSION

Embodiments and features discussed above can be realized in the form of information stored in volatile or non-volatile computer or device readable media. This is deemed to include at least media such as optical storage (e.g., compact-disk read-only memory (CD-ROM)), magnetic media, flash read-only memory (ROM), or any current or future means of storing digital information. The stored information can be in the form of machine executable instructions (e.g., compiled executable binary code), source code, bytecode, or any other information that can be used to enable or configure computing devices to perform the various embodiments discussed above. This is also deemed to include at least volatile memory such as random-access memory (RAM) and/or virtual memory storing information such as central processing unit (CPU) instructions during execution of a program carrying out an embodiment, as well as non-volatile media storing information that allows a program or executable to be loaded and executed. The embodiments and features can be performed on any type of computing device, including portable devices, workstations, servers, mobile wireless devices, and so on.

Claims

The invention claimed is:

1. A computer-implemented method of updating a virtual machine, the method comprising:

executing a hypervisor on a host computer, the hypervisor managing execution of the virtual machine on the host computer, wherein a guest operating system is configured to execute on the virtual machine, the hypervisor configured to provide the virtual machine with block-based access to a virtual machine image, the virtual machine image comprised of blocks;

receiving a substitution block at the host computer; and

while the virtual machine is live, replacing, by the hypervisor, a target block in the virtual machine image with the substitution block, the target block comprising one of the blocks of the virtual machine image, the hypervisor managing the blocks of the virtual machine image and, when the virtual machine is executing, presenting the blocks of the virtual machine image, including the substitution block, to the virtual machine as a virtualized disk.

2. A method according to claim 1, further comprising receiving a list of blocks, and using the list of blocks to identify the target block.

3. A method according to claim 1, wherein the substitution block was obtained from a base virtual machine image in relation to application of a patch thereof, and wherein the virtual machine image was derived from the base virtual machine image before the patch was applied.

4. A method according to claim 1, wherein the target block contains a first version of a portion of a binary machine code file executable by the guest operating system, the substitution block comprises a second version of the binary machine code file, and when the substitution block replaces the target lock in the virtual machine image, the second version of the binary machine executable file is available for execution by the guest operating system.

5. A method according to claim 1, wherein the virtual machine is paused or suspended during the replacing.

6. A method according to claim 1, wherein the virtual machine is executing during the replacing.

7. A method according to claim 1, the method further comprising providing, by the hypervisor, a signal to the guest operating system indicating the replacing, and wherein the guest operating system responds to the signal by executing a corresponding action.

8. One or more computer-readable storage devices storing information to enable a computer to perform a process, the process performed by a virtualization layer executing on the computer, the process comprising:

executing the virtualization layer on the computer, the virtualization layer configured to manage execution of a virtual machine on the computer, wherein a guest operating system is configured to execute on the virtual machine, the virtualization layer configured to provide the virtual machine with block-based access to a first virtual machine image if the virtual machine is executing, the first virtual machine image comprised of first blocks that are units of storage that the virtualization layer presents as a virtualized disk;

receiving a second block obtained from a second virtual machine image; and

while the virtual machine is live, replacing one of the first blocks with the second block.

9. One or more computer-readable storage devices according to claim 8, wherein the virtual machine executes from the virtualized disk and the virtualized disk comprises a differencing disk, and wherein the second block is added to the differencing disk without removing the replaced first block from the virtualized disk.

10. One or more computer-readable storage devices according to claim 9, wherein the replacing is performed such that the process is specific to the guest operating system.

11. One or more computer-readable storage devices according to claim 8, wherein the hypervisor provides the guest operating system with information identifying the replaced first block.

12. One or more computer-readable storage devices according to claim 11, wherein the guest operating system uses the information identifying the replaced block to reload a corresponding unit of storage managed by the guest operating system.

13. One or more computer-readable storage devices according to claim 8, wherein the virtual machine is suspended or paused by the hypervisor during the replacing.

14. A method performed by one or more computers, the method comprising:

modifying a first virtual machine image to generate a second virtual machine image;

identifying blocks of the first virtual machine image that were modified by the modifying;

distributing the identified blocks to a host, the host having a hypervisor and one or more virtual machines managed thereby, the virtual machine having a virtual machine image comprised of blocks and presented as a virtual disk by the hypervisor; and

while the hypervisor keeps the virtual machine in a live state, applying the blocks to the virtual machine on the host, the applying performed by the hypervisor replacing blocks of the virtual machine image managed by the hypervisor.

15. A method according to claim 14, the process further comprising pausing or suspending the virtual machine prior to the applying and resuming execution of the virtual machine after the applying.

16. A method according to claim 15, further comprising responding, to an indication that the applying was performed, by the guest operating system using the indication to identify storage elements managed by the guest operating system that are to be re-read by the guest operating system.

17. A method according to claim 15, further comprising using a file-to-block mapping on the host to identify a file managed by the guest operating system.

18. A method according to claim 17, further comprising forcing a page fault to force executable code in one of the blocks into memory of the guest operating system.

19. A method according to claim 15, wherein the virtual machine is executing when the replacing is performed.

20. A method according to claim 15, wherein the hypervisor comprises a block-based device driver that handles access to a virtual machine image of the virtual machine by mapping blocks of the virtual machine image to a virtual disk of the virtual machine, the block-based device driver providing the applied blocks to the virtual machine.